Skip to content

Snowflake

Coming Soon

Snowflake support is under active development. This page describes the planned integration.

Overview

Snowflake is a cloud-native data warehouse with automatic scaling, columnar storage, and built-in query result caching. It's an ideal backend for GradientHarbor analytics workloads.

┌─────────────────┐     ┌──────────────────┐     ┌─────────────────┐
│ GradientHarbor  │────▶│   Snowflake      │────▶│  Cloud Storage  │
│                 │ SQL │                  │     │  (S3/GCS/Azure) │
│  AI Agent       │◀────│  Query Engine    │◀────│  Columnar Data  │
│  Dashboards     │     │  (auto-scaling)  │     │                 │
└─────────────────┘     └──────────────────┘     └─────────────────┘

Planned Configuration

FieldDescription
AccountYour Snowflake account identifier (e.g., xy12345.us-east-1)
WarehouseCompute warehouse to use for queries
DatabaseDefault database for catalog introspection
UsernameSnowflake user for authentication
PasswordSnowflake password

Why Snowflake

  • Columnar storage — Optimized for the analytical queries GradientHarbor generates
  • Auto-scaling — Warehouses scale up/down based on query complexity
  • 24-hour query cache — Repeated queries return instantly at zero compute cost
  • Multi-cluster — Concurrent dashboards don't compete for resources
  • Time travel — Query historical data snapshots

Best Practices (Planned)

  • Use a dedicated warehouse for GradientHarbor queries to isolate BI workloads
  • Set warehouse auto-suspend to minimize idle costs
  • Use X-Small warehouse size for most dashboard queries (sufficient for aggregations)
  • Enable query result caching (on by default) for maximum benefit from GradientHarbor's three-tier cache